Customer insurance marketing application management system

Abstract

The invention discloses an insurance marketing application management system, which realizes real-time data updating and personalized guidance suggestions of each role person under different contacts through data and information circulation under an insurance marketing full-service scene. Therefore, related personnel can know own data change and the next action direction at any time and receive electronic knowledge manual related to guidance suggestions. The decision and operation of each role personnel can readjust the layout and optimization of different contacts and next decision points in the whole business process chain to form an optimal network node system, so that the real-time circulation of data and information is realized, and each role personnel is ensured to be always in the latest information environment. Along with the proceeding of each operation and decision, the system can dynamically adjust suggestions and guidance of the next step according to the characteristics of personal roles and business processes, so that the user is helped to complete tasks more efficiently.

Description

Technical Field

[0001] The present invention relates to the technical field of information service systems, and in particular to a policyholder marketing application management system. Background Art

[0002] Insurance marketing is a strategy for enterprises to mine sales leads for existing customers, aiming to encourage these customers to increase or replace the company's products in order to achieve sales targets. However, in the current market environment, most companies still use traditional methods for insurance marketing, which has many problems: single means, difficulty in accurately grasping customer needs, low conversion rate, etc. In addition, marketing activities are usually led by sales staff, who take sales promotion as their primary goal, and customer experience is often ignored. At the same time, employees from other departments are often excluded from insurance marketing, such as after-sales service staff, customer service staff, and second-hand goods department staff, who are more familiar with customers' current needs and usage. Employees in these departments may have more information about customers' purchasing needs, but this information has not been fully shared and utilized. Therefore, problems such as information asymmetry and poor communication between departments directly affect the effectiveness of insurance marketing. At present, the promotion of insurance marketing business of enterprises mostly relies on personnel promotion, and there are few auxiliary tools, models or systems to improve accuracy and work efficiency.

[0003] In the absence of digital empowerment for insurance customer marketing, the automotive industry faces a series of serious drawbacks and defects.

[0004] First, due to the lack of scientific system and tool support, business processes rely on human efforts, which leads to low efficiency, increased costs and unstable work quality. The inability to achieve automated and intelligent process management makes it difficult for companies to cope with rapidly changing market demands.

[0005] Secondly, even if there is system support, it is limited to simple static storage of data and lacks real-time update and iteration mechanisms, resulting in the inability to effectively guarantee the timeliness of data. This will lead to untimely and inaccurate information, inability to quickly respond to customer needs, and missed business opportunities.

[0006] Third, the lack of data sharing and collaboration mechanisms has led to information silos between multiple parties involved in the business, making it impossible to communicate and collaborate effectively. This not only affects the accuracy of decision-making, but also leads to a decline in customer experience and ultimately affects sales performance.

[0007] In summary, insurance marketing without digital empowerment will face a series of problems such as low efficiency, inaccurate information, and loss of business opportunities, which will affect the company's competitiveness and sustainable development. Therefore, digital transformation is an inevitable trend in the development of insurance marketing. Only through the empowerment of technology can we achieve continuous innovation in business and improve performance. Summary of the invention

[0008] In view of the above-mentioned defects in the prior art, the present invention proposes a policyholder marketing application management system, which transforms the marketing model guided by subjective experience into a quantifiable and practical standard process through a scientific algorithm, and realizes the methodology and implementation method of online full-closed-loop information flow between multiple roles at work.

[0009] In order to achieve the above object, the present invention provides the following technical solutions:

[0010] The present invention provides a policyholder marketing application management system, which includes the following modules:

[0011] Data collection and storage module: used to collect data from different user roles in different business scenarios and store it in the central database;

[0012] Data processing and analysis module: Analyzes the collected data through machine learning algorithms, matches the decision paths in the central database, and sends the optimal decision path to the guidance suggestion generation module;

[0013] Guidance and suggestion generation module: Based on the data analysis results and user roles, the system generates corresponding guidance and action suggestions;

[0014] Data flow and push module: pushes the generated guidance suggestions to relevant role users and ensures the flow of information at different touchpoints;

[0015] Knowledge manual matching module: matches and pushes relevant electronic knowledge manuals based on guidance suggestions and user needs;

[0016] User feedback and update module: used to collect feedback data after users receive guidance suggestions and perform corresponding operations, and update the database to optimize the next guidance suggestions.

[0017] Furthermore, the user roles include general manager, sales director, service director, marketing director, used car director, financial director, sales manager, service manager, customer service manager, ADC manager, sales consultant, service consultant, used car consultant, used car appraiser, marketing specialist, insurance specialist, customer service specialist, and ADC specialist.

[0018] Furthermore, the business scenarios include top-level scenarios, parent-level scenarios and child-level scenarios, wherein the top-level scenarios include sales scenarios and after-sales and service scenarios, the parent-level scenarios include new car sales, used car sales, after-sales service, customer car use, and customer churn scenarios, and the child-level scenarios include after-sales maintenance in-store, second showrooms, customer return visits, event invitations, financial renewals, and additional purchase scenarios.

[0019] Furthermore, the data processing and analysis module adopts a decision tree regression model, so that each decision node is the optimal decision, and the subsequent decisions are re-coordinated according to the decision of the previous node, so that the entire decision-making system is always in the optimal decision.

[0020] Furthermore, the calculation process of the decision tree regression model is as follows: if the customer currently served by the user is in a certain scenario, the system will give priority to recommending according to a certain first-level decision; when the user selects and executes the first-level decision, the system will recommend a second-level decision. If the user does not choose according to the optimal recommendation in the first step, but chooses another first-level decision, the system will calculate and determine the subsequent optimal decision point based on the current decision point; and so on, gradually completing the optimal decision of each decision node until the insurance customer marketing service is completed;

[0021] The above recommendation algorithm adopts the optimized equal probability criterion method. First, each decision node is quantitatively assigned a value according to the actual business value, and then the value score of each action plan at each node is calculated. Then, the results after the implementation of each action plan are compared, and the action with the maximum value is taken as the decision basis for the optimal action. The optimal decision value Q(s,a) of the optimal decision route is opt ) is calculated as:

[0022] Q(s,a opt )=Max{(1 / n)∑u ij (s i ,a j )}

[0023] in:

[0024] s i : a certain node;

[0025] a j : a decision-making route;

[0026] u ij : The value score of a node on a route;

[0027] n: the number of nodes on a decision path.

[0028] Compared with the prior art, the present invention has the following beneficial effects:

[0029] The policyholder marketing application management system provided by the present invention has the advantage that it can solve the various problems faced by traditional policyholder marketing, can provide professional decision-making support for inexperienced employees, can realize real-time information sharing and data interaction during the decision-making execution process, and can incorporate relevant role personnel into a closed-loop process system to build a dynamic, interactive, and analyzable digital system, thereby improving business efficiency and customer experience.

[0030] This invention successfully overcomes a major problem faced by traditional insurance marketing, that is, transforming the original sensory, qualitative and unstructured workflow into a quantitative workflow guided by scientific experience and derived through precise algorithms. This measure greatly compensates for the work deficiencies caused by human limitations, thereby effectively improving work efficiency and quality. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0032] Figure 1 This is a schematic diagram of the decision tree regression model provided by the present invention.

[0033] Figure 2 This is a task distribution and flow chart provided by the present invention.

[0034] Figure 3 The user role classification provided by the present invention.

[0035] Figure 4 This is a schematic diagram of role transfer provided by the present invention. DETAILED DESCRIPTION

[0036] In order to better understand the technical solution, the method of the present invention is described in detail below with reference to the accompanying drawings.

[0037] The main business system of the present invention uses the WeChat ecosystem as the development environment. The main business system develops WeChat applets and WeChat service accounts, and the management background is developed using the computer-side PHP language. The present invention adopts the SaaS software service model, which is mainly suitable for the insurance customer marketing demand scenarios in the automotive industry. The mobile terminal uses the applet technology products and WeChat push services to efficiently provide dealers with insurance customer marketing training, information query, insurance customer marketing customer information management follow-up and other services. Provide strong support for the entire life cycle of the software system, including system design, development and deployment, operation and maintenance, and version updates! The programming language is JavaScript, which has strong cross-platform capabilities and can be published to different platforms.

[0038] The core technical solution of the present invention is to realize real-time data update and personalized guidance suggestions for each role at different touch points through the data and information flow in the whole business scenario of insurance marketing. This allows relevant personnel to understand their data changes and next action direction at any time, and receive electronic knowledge manuals related to guidance suggestions. The decisions and operations of each role will readjust the layout and optimization of different touch points and next decision points in the entire business process chain, forming an optimal network node system.

[0039] The key to this optimization solution is to achieve real-time data and information flow, ensuring that personnel in various roles are always in the latest information environment. With each operation and decision, the system will dynamically adjust the next step of suggestions and guidance based on the characteristics of individual roles and business processes, helping users complete tasks more efficiently.

[0040] In addition, the optimization plan also includes pushing electronic knowledge manuals, so that users can obtain relevant knowledge and guidance at any time when performing tasks, improving work quality and efficiency. Such personalized guidance and knowledge support not only improves the execution efficiency of the entire business process, but also enhances the professional ability and job satisfaction of employees.

[0041] Overall, through the optimized core technology solution, the digital system of Baoke Marketing has achieved more intelligent and personalized data flow and guidance suggestions, provided the optimal network node system for each operation and decision, and improved the efficiency and quality of business processes.

[0042] The insurance customer marketing application management system provided by the present invention includes a data collection and storage module, a data processing and analysis module, a guidance and suggestion generation module, a data flow and push module, a knowledge manual matching module and a user feedback and update module. Among them:

[0043] 1. Data acquisition and storage module

[0044] This module is used to collect data from different user roles in different business scenarios and store it in a central database.

[0045] The user roles include general manager, sales director, service director, marketing director, used car director, financial director, sales manager, service manager, customer service manager, ADC manager, sales consultant, service consultant, used car consultant, used car appraiser, marketing specialist, insurance specialist, customer service specialist, ADC specialist, etc. The key information they collect during the reception or communication and negotiation with customers will be entered and stored in the central database through this patented system.

[0046] The business scenarios include top-level scenarios, parent-level scenarios and child-level scenarios, among which the top-level scenarios include sales scenarios and after-sales and service scenarios, the parent-level scenarios include new car sales, used car sales, after-sales service, customer car use, customer churn and other scenarios, and the child-level scenarios include after-sales maintenance in-store, second exhibition hall, customer return visits, event invitations, financial renewal, additional purchases and other scenarios.

[0047] 2. Data processing and analysis module

[0048] This module analyzes the collected data through machine learning algorithms, matches the decision paths in the central database, iteratively updates the knowledge base content through existing data and operation records of users with different roles, and cyclically optimizes the data training and update process to provide more accurate and effective guidance and suggestions, and send the optimal decision path to the guidance and suggestion generation module.

[0049] The data processing and analysis module adopts a decision tree regression model, so that each decision node is the optimal decision, and the subsequent decisions are re-coordinated according to the decision of the previous node, so that the entire decision-making system is always in the optimal decision.

[0050] The decision tree regression model calculation process is as follows Figure 1 As shown, specifically: if the customer currently served by the user is in scenario 1, the system will give priority to recommending according to the first-level decision 1-2; when the user selects and executes the first-level decision 1-2, the system will recommend the second-level decision 2-7. If the user does not choose according to the optimal recommendation in the first step, but chooses the first-level decision 1-3, the system will calculate and determine the subsequent optimal decision point based on the current decision point; and so on, gradually completing the optimal decision of each decision node until the policyholder marketing service is completed.

[0051] The above recommendation algorithm adopts the optimized equal probability criterion method. First, each decision node is quantitatively assigned a value based on the actual business value. The actual business value comes from the training accumulation and rolling update of historical practice data. The quantitative assignment is based on the expert's experience score and is updated to the system as the input parameter of the model. Then, the value score of each action plan (i.e., decision route) at each node is calculated. The value score is calculated using the parameters and formulas in the following optimal decision model. Then, the results after the implementation of each action plan are compared, and the action with the maximum value is taken as the decision basis for the optimal action; the optimal decision value Q(s,a) of the optimal decision route opt ) is calculated as:

[0052] Q(s,a opt )=Max{(1 / n)∑u ij (s i ,a j )}

[0053] in:

[0054] s i : a certain node;

[0055] a j : a decision-making route;

[0056] u ij : The value score of a node on a route;

[0057] n: the number of nodes on a decision path.

[0058] For example, when role a receives a customer, after initially understanding certain customer needs in the first stage, the system will match the decision path in the central database based on the needs, and traverse and calculate all the paths in the central database, and finally find an optimal decision path and recommend it to role a for selection; when the next step of communication with the customer is carried out based on the decision suggestion after the selection, the decision path in the central database will be matched again based on the new communication needs, and finally an optimal decision path will be found and recommended to role a for selection, and so on, finally completing the complete communication with the customer and finally facilitating the transaction.

[0059] As shown in Table 1: According to the value score calculation of different decision nodes, decision line a4 is the optimal decision strategy.

[0060] Table 1

[0061] <![CDATA[u ij ]]> <![CDATA[s1]]> <![CDATA[s2]]> <![CDATA[s3]]> <![CDATA[s4]]> <![CDATA[s5]]> 1 / 5 max <![CDATA[a1]]> 0 0 0 0 0 <![CDATA[a2]]> -10 20 20 20 20 14 <![CDATA[a3]]> -20 10 40 40 40 22 <![CDATA[a4]]> 30 0 -30 60 60 24 24 <![CDATA[a5]]> -40 20 -10 50 80 20

[0062] 3. Guidance suggestion generation module

[0063] Based on the data analysis results (behavioral history (user operations and information records in the mini program) and the current customer's application scenario (such as maintenance in the store, invitation to participate in marketing activities, etc.)) and user roles, the system generates corresponding guidance suggestions and action suggestions, fully considering user preferences and business requirements.

[0064] Specifically, according to the data processing and analysis of the data processing and analysis module, each decision node is an analysis result, and the guidance suggestions under the analysis result are the current optimal decision, and guide the system users to take appropriate action plans. The analysis results, guidance suggestions and action plans are all carried and operated through the WeChat applet.

[0065] 4. Data transfer and push module

[0066] This module pushes the generated guidance suggestions to relevant role users and ensures the flow of information at different touchpoints. For example, guidance suggestions are pushed to role personnel related to the person who will execute the suggestion through page presentation or WeChat instant messaging. The push recipient can directly enter the mini program in WeChat instant messaging to perform operations after the flow. Specifically, the data flow path and rules are designed to ensure the accurate transmission and update of information at different roles and touchpoints, such as Figure 2 As shown in the figure, the implementation process of the data flow algorithm is as follows:

[0067] 1) Role A receives a customer for the first time, creates a customer profile, and assigns tasks to Role B;

[0068] 2) Role B participates in the customer's communication activities based on the customer information and communication records recorded by Role A, and retains the necessary communication information. If Role C is required to participate in the customer's service process, a task is created and distributed to Role C;

[0069] 3) After role C joins, he can share the communication records between roles A and B to provide background reference for his own customer service;

[0070] 4) And so on, until the service to the customer is terminated, thus achieving the goal of policyholder marketing.

[0071] Through the data flow and push module, Insurance Marketing can achieve real-time information sharing and data interaction. There are no more information islands and data fragmentation problems. Data can be shared between departments and personnel to achieve smooth information transmission and instant updates. This makes decision-making more accurate and rapid, and also improves teamwork efficiency. The digital system incorporates all roles and personnel related to the Insurance Marketing business into a closed-loop process system. Each link can be recorded, analyzed and optimized to form a complete business closed loop. In this way, the entire process becomes more transparent and controllable, and problems and bottlenecks can be quickly identified and resolved.

[0072] In addition, the data flow and push module can also provide interactive and analytical functions. Users can obtain business data, reports and analysis results at any time as needed, and conduct real-time monitoring and decision-making. This enables managers to better understand business conditions and adjust strategies in a timely manner to respond to market changes and customer needs, and achieve information synchronization, data interaction and sharing between "people-things-objects" across time and space.

[0073] 5. Knowledge manual matching module

[0074] This module matches and pushes relevant electronic knowledge manuals based on the guidance suggestions and user needs, providing users with more in-depth support and guidance. Specifically, the central database stores the electronic knowledge manual knowledge base, and each guidance suggestion establishes a corresponding relationship with the electronic manual in the knowledge base. When a specific suggestion is determined at a certain node in the circulation process, the corresponding knowledge manual can be automatically matched in the knowledge base.

[0075] User feedback and update module: used to collect feedback data after users receive guidance suggestions and perform corresponding operations, and update the database to optimize the next guidance suggestions.

[0076] This module uses the data collection and storage module to track and record the user's execution status, including the user's input data, execution results, and whether the conditions are met. The results of model training update the guidance suggestions and data in the database. Finally, the updated database is applied to the next guidance suggestions, and the feedback data of user execution is continuously collected to cyclically optimize the data training and update process to provide more personalized and effective guidance suggestions.

[0077] In summary, the policyholder marketing application management system provided by the present invention improves the efficiency of information sharing and data interaction, realizes closed-loop process management, and provides interactive and analyzable functions, thereby helping enterprises better cope with competitive pressure and improve business level and customer satisfaction.

[0078] Application Example 1

[0079] 1. Create login accounts for different roles (such as Figure 3 As shown in the figure, different data permissions and process permissions are set for different login accounts based on their roles (e.g., consultant-level accounts can only view and maintain the customer information they are responsible for, manager-level accounts can view the information and data corresponding to the consultant accounts under their management, and director-level accounts can view the information and data under the management of all managers). For example, different role accounts may experience different business process nodes, and the page content or operation steps they can see may also be different.

[0080] 2. According to different scenarios of communication with customers, the system sets different application scenario entrances. For example, when a car owner needs to return to the store for maintenance, the service consultant who receives the customer finds out during the communication that the user plans to change the car. Then the service consultant can enter from the "Maintenance in the store" entrance and create a customer file. At the same time, if it is found that the user plans to sell the existing used car, then he will invite a used car consultant to join the user's communication chain. The used car consultant can see all the customer's information and the current stage of communication with the customer through the flow of system data, and will know what to do for the customer in the current stage.

[0081] 3. After receiving relevant instructions, the personnel in the next link will quickly browse the work records retained by all participating members in the customer chain, and communicate with the customer as soon as possible to complete the work that needs to be done in the current link.

[0082] 4. When all the nodes in the chain are completed, the insurance marketing work for the customer is all over. Regardless of whether the customer is successfully converted into a repeat customer, the customer will be closed and archived. This work is completed by the last participating member in the chain.

[0083] 5. Accounts with management roles can view and analyze the data information generated during the entire process and provide decision-making guidance.

[0084] Application Example 2

[0085] Car owner A comes to the store for repair and maintenance. Service consultant A is responsible for receiving the customer. During the communication, it is found that customer A intends to exchange the car. The operation path of service consultant A is:

[0086] 1. Service consultant A logs in to the mini program with his consultant account and enters the "After-sales maintenance in store" scenario page. The system matches the strategy in the central database based on the "After-sales maintenance in store" scenario, prompts consultant A and displays the corresponding communication script on the page, asking whether he plans to change the car;

[0087] 2. When the user answers "yes", service consultant A records customer A's needs on the flow tool page. At the same time, at the current node, the system provides at least three decision paths through decision judgment (the judgment method refers to the decision model in the solution details section):

[0088] (1) Invite a sales consultant to participate in introducing the new car to the customer (according to the central database maintenance, this path decision is divided into 4 points);

[0089] (2) Invite used car consultant B to participate in the residual value assessment of the used car (this path decision is scored 5 points);

[0090] (3) Invite financial advisor C to participate and provide him with a financial plan for a loan to buy a car (this path decision is scored 3 points);

[0091] 3. According to the optimal decision principle, the system prompts that decision 2 is the best recommendation. At this time, service consultant A adds used car consultant B to the process in the circulation tool;

[0092] 4. Used car consultant B receives a timely message from WeChat and clicks on it to enter the mini program to view the current demand and subsequent processes. At this time, the system prompts used car consultant B to proactively contact the customer to conduct a used car residual value assessment. At the same time, the central knowledge base provides communication scripts;

[0093] 5. Used car consultant B records the used car evaluation report in the process file of customer A through the circulation tool. Make a decision based on the residual value, vehicle condition, mileage, year of purchase and other data in the report (refer to the decision model in the solution details for the judgment method), and provide at least 2 decision paths:

[0094] (1) Help customer A develop a second-hand car replacement plan and a financial loan plan for purchasing a car (this path decision is divided into 4 points), and invite financial advisor C to participate;

[0095] (2) Helping Client A develop a second-hand car replacement plan and a financial loan plan for purchasing a car (this path decision is divided into 2 points). Financial advisor C is not invited to participate at this time;

[0096] 6. After seeing the replacement and financial plan recorded by the financial advisor through the circulation tool, service consultant A has in-depth communication with customer A. At this time, the system retrieves the communication script template from the central knowledge base and presents it to service consultant A through the mini program;

[0097] 7. If customer A accepts this plan, the service consultant will make a decision (refer to the decision model in the plan details section for the judgment method) and prompt that the sales consultant should be invited to participate. At this time, service consultant A will add sales consultant D to the communication process with this customer.

[0098] 8. The system notifies sales consultant D via WeChat instant messaging. Sales consultant D enters the mini program to view the files and communication information of past customer A and communicates with customer A about the sales process. At this time, the system retrieves the sales communication script template from the central knowledge base and presents it to sales consultant D through the mini program.

[0099] 9. Customer A agrees to purchase, and Sales Consultant D records the information such as the intended model, purchase price, and purchase date of Customer A in the circulation tool;

[0100] 10. Customer A disagrees to purchase, and Sales Consultant D records the reason for disagreement in the circulation tool. At this time, the system retrieves the defeat communication script template from the central knowledge base and presents it to Sales Consultant D through the mini program. At the same time, Service Consultant A maintains follow-up communication with Customer A.

[0101] Role flow Figure 4 shown.

[0102] Compared with traditional insurance marketing, the solution of the present invention has the following beneficial effects:

[0103] Process optimization: Through the generation of personalized suggestions and optimization of data flow, the policyholder marketing process can be simplified, redundant steps and information transmission links can be reduced, and the overall process efficiency can be improved.

[0104] Improved efficiency: The system can automatically generate guidance suggestions based on user roles and behavior patterns, saving time in manually making suggestions and improving operational efficiency and response speed.

[0105] Improved conversion rate: Personalized recommendations can more accurately collect user needs, improve user experience and satisfaction, and thus increase user conversion rate and loyalty.

[0106] In order to compare the effects of the traditional insurance customer marketing scheme and the scheme of the present invention, the following test evaluation method can be used:

[0107] A / B testing: Some users are given the traditional policyholder marketing program, while others are given the new program. The conversion rates, satisfaction, and efficiency indicators of the two groups of users are compared (as shown in Table 2) to evaluate the pros and cons of the new program.

[0108] User feedback survey: Collect user feedback and experience on traditional and new policyholder marketing programs, understand users' views on both, and thus evaluate the actual effect of the new program.

[0109] Data analysis and comparison: Through data analysis of the implementation of traditional and new solutions, we compare key indicators such as process efficiency and conversion rate between the two to verify the advantages of the new solution.

[0110] Table 2

[0111]

[0112] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, a person skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features thereof may be replaced by equivalents. However, such modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A policyholder marketing application management system, characterized in that: Includes the following modules: Data collection and storage module: used to collect data from different user roles in different business scenarios and store it in the central database; Data processing and analysis module: Analyzes the collected data through machine learning algorithms, matches the decision paths in the central database, and sends the optimal decision path to the guidance suggestion generation module; Guidance and suggestion generation module: Based on the data analysis results and user roles, the system generates corresponding guidance and action suggestions; Data flow and push module: pushes the generated guidance suggestions to relevant role users and ensures the flow of information at different touchpoints; Knowledge manual matching module: matches and pushes relevant electronic knowledge manuals based on guidance suggestions and user needs; User feedback and update module: used to collect feedback data after users receive guidance suggestions and perform corresponding operations, and update the database to optimize the next guidance suggestions.

2. The policyholder marketing application management system according to claim 1, characterized in that: The user roles include general manager, sales director, service director, marketing director, used car director, financial director, sales manager, service manager, customer service manager, ADC manager, sales consultant, service consultant, used car consultant, used car appraiser, marketing specialist, insurance specialist, customer service specialist, and ADC specialist.

3. The policyholder marketing application management system according to claim 1, characterized in that: The business scenarios include top-level scenarios, parent-level scenarios and child-level scenarios, among which the top-level scenarios include sales scenarios and after-sales and service scenarios, the parent-level scenarios include new car sales, used car sales, after-sales service, customer car use, and customer churn scenarios, and the child-level scenarios include after-sales maintenance in-store, second showrooms, customer return visits, event invitations, financial renewals, and additional repurchase scenarios.

4. The policyholder marketing application management system according to claim 1, characterized in that: The data processing and analysis module adopts a decision tree regression model, so that each decision node is the optimal decision, and the subsequent decisions are re-coordinated according to the decision of the previous node, so that the entire decision-making system is always in the optimal decision.

5. The policyholder marketing application management system according to claim 4, characterized in that: The calculation process of the decision tree regression model is as follows: if the customer currently served by the user is in a certain scenario, the system will give priority to recommending according to a certain first-level decision; when the user selects and executes the first-level decision, the system will recommend a second-level decision. If the user does not choose according to the optimal recommendation in the first step, but chooses another first-level decision, the system will calculate and determine the subsequent optimal decision point based on the current decision point; and so on, gradually completing the optimal decision of each decision node until the insurance customer marketing service is completed; The above recommendation algorithm adopts the optimized equal probability criterion method. First, each decision node is quantitatively assigned a value according to the actual business value, and then the value score of each action plan at each node is calculated. Then, the results after the implementation of each action plan are compared, and the action with the maximum value is taken as the decision basis for the optimal action. The optimal decision value Q(s,a) of the optimal decision route is opt ) is calculated as: Q(s,a opt )=Max{(1 / n)∑u ij (s i ,a j )} in: s i : a certain node; a j : a decision-making route; u ij : The value score of a node on a route; n: the number of nodes on a decision path.

Images